The big challenge clinicians face is translating clinical trial data into a treatment for individual patients. For a younger patient with multiple lesions in the brainstem and the spinal cord, the prognosis is relatively poor. However, while we can readily assess patients as having good or bad prognostic factors, we are less good at predicting which individual patients will have a favourable disease course. Equally, we cannot say with any accuracy, which is the right drug for an individual patient. It is a key concept that nobody can predict individual treatment response at baseline.
Weighing risk and benefit may be completely different to the patient than the physician. Patients are less concerned about risk of PML than physicians. This is probably because physicians have experience of PML developing and causing mortality, whereas patients most likely have not.